Anonymous interactions within the metagovernment
There is an ongoing debate in the Metagovernment community over the place of anonymous participation.
The question may have to be treated separately for different kinds of actions, such as participating in discussions, editing or making proposals, and voting. The issue of anonymous voting has provoked the most debate and is currently the focus of this article.
It should also be noted that different communities adopting collaborative governance tools may wish to make their own decision on how to treat anonymous participation.
- 1 Anonymous voting or not? Possible solutions
- 2 Arguments for allowing anonymous voting
- 3 Arguments against anonymous voting
- 4 References
Anonymous voting or not? Possible solutions
There are numerous possible solutions to the question of whether and how to allow anonymous voting. The following is a non-exhaustive list:
- All users must be authenticated as real persons and all voting must be public
- All users must be authenticated as real persons, but authenticated users can choose in each case for their vote to remain anonymous (for example, using public-key cryptography).
- Users can create non-identifying personas, so personas are authenticated within the community via a trust-metric, but no one has to know who they are in the real world
- Anonymous users are allowed to vote and these votes are given equal weight
- Anonymous users are allowed to vote, but these votes are given less weight than those of authenticated user
Arguments for allowing anonymous voting
The problem of coercion is that if there's a way of finding out how a certain person voted, one can apply pressure to get that person to vote a particular way. For example, an employer may make an implicit or explicit threat to fire employees who don't take the company line on a certain vote; an organised crime gang might make a threat of violence to a community; a political organisation might expel members who vote in line with the organisation; a telecommunications provider might be slower to respond to technical breakdowns of customers who voted against its planned construction project etc.
Preventing vote buying
Whereas coercion can be expressed as, "Do it my way or something bad happens", it can also easily be reversed as, "Do it my way and something good happens", as in the case of vote buying.
Social pressure / self-censorship
Individuals may feel pressured to conform to the norms of their family, community, church, corporation etc. and alter their voting behaviour accordingly. This behaviour, which we may also call self-censorship, is regarded as a corruption of public voting since people are not voting the way they want to. However, it may be impossible to identify as there is no explicit coercion involved.
Protecting public reputation
Web-based public actions are permanent - something recorded once is often archived and mirrored such that it can never be repressed. This presents additional ramifications for individuals. For example, someone trying to apply for a job may have their potential future employer look into their voting history.
Assuming some people wish to vote anonymously, they should have the facility to do so regardless of their reasons. Otherwise the results of a vote may be skewed as certain people choose not to participate.
Arguments against anonymous voting
Vote buying would be uneconomical
In order to make a real difference to a vote, the person/organisation making the threat or bribe would have to reach out to a significant number of people. In the case of vote buying, this may simply prove uneconomical or draw too much attention to the coercion. This argument is strengthened in a situation where there are a large number of votes on small issues, rather than a few votes on big issues.
Coercion would be counterproductive
Just as, for instance, an employer can verify, in a public vote, whether its employees have voted in a certain way (on threat of dismissal), so too could the public subsequently verify that an unusually large number of employees had voted 'against' their employer and been dismissed. Alternatively, all the employees could vote as the company wishes, but someone could anonymously leak evidence of the coercion, such as an internal email, into the public domain. As such, organisations/individuals attempting to influence a vote risk being caught out, resulting in a) a wider public voting a different way on the relevant issue, b) damage to the reputation of the person/organisation trying to influence the vote c) in case of a breach of the law, prosecution.
Further, although compliance with a bribe may easily be verified in a public vote, there is no guarantee of continued compliance. The voter may take the money from one side, then shift her vote and take it from the other.
Tweaking towards a consensus
Online voting systems need not follow the pattern of large, controversial yes/no votes. We can allow each voter to have his own unique position on any subject. In a vote regarding whether to legalise marajuana, for instance, the voter must not choose between yes or no, instead he can have the position "I want further scientific research into the potential benefits of marijuana." In this way, when it comes to such delicate questions, a whole society can move slowly and gently over time towards its consensus. In other words, by tweaks rather than by reforms.
Further, the fact that we are faced with a much larger number of small choices on an infinite number of issues may lead to a greater tolerance between people in their various social circles, as it would be impossible to please everyone all the time even if that was your only aim.
Coercion can still take place with private voting
Even if people know their votes cannot be checked by a third party, they may still be coerced. For instance, the claim of a threat of terrorist violence may be publicly exaggerated in order to promote a vote in favour of an aggressive defense policy.
This argument states that if coercion is a problem, then it is strictly a social problem. If the private sphere of individuals, families, employers, and so forth, is restricting the public communications of individuals wrongly, in defiance of the norms, then society itself has a problem in the relations between its private and public spheres. It is not a problem for a voting medium that functions exclusively in the public sphere.
Ensuring anonymity may be impossible
Given a complex voting system which allows a large number of voting combinations, it may be possible to verify whether someone has voted the way they were asked to - even if the voting is private - by checking if that exact voting combination is present. For instance, when voters are asked to list 10 options in order of preference, someone trying to corrupt the vote could ask a large number of people to each vote for a specific combination, all of which have the desired winning option placed first or second.
A more resistent system
An entirely public voting system may be more resistent to manipulation, as it would be much easier to detect unusual voting patterns which may point to cases of manipulation.
Metagovernment Start Archives: *http://metagovernment.org/pipermail/start_metagovernment.org/